smart maintenance in cyber- physical...

19
Smart Maintenance in Cyber- Physical Systems Harald Rødseth Postdoctoral Fellow Department of Mechanical and Industrial Engineering, NTNU

Upload: dangdung

Post on 27-Jul-2018

214 views

Category:

Documents


0 download

TRANSCRIPT

Smart Maintenance in Cyber-Physical Systems

Harald RødsethPostdoctoral FellowDepartment of Mechanical and Industrial Engineering, NTNU

2

Aim and outline

Aim of this presentation: To present concepts that enable Smart Maintenance in Cyber-physical systems

Outline:

1. Cyber-physical systems

2. Smart Maintenance

3. Application of Smart maintenance in cyber-physical

systems

4. Summary & Conclusion

3

1.Cyber-physical systems

1838: SS Great western 1911: Selandia

1937: 167 men for operating an engine room!

1961: M/S Kinkasan Maru

1966: DNV E02012: Concept project for deep sea vessel

Age of sail

Steam engine Diesel engine

Instrumentation & automation of Engines

MUNIN project

1969: M/S Taimyr

4

1.Cyber-physical systems

Figure 1: Four stages of industrial revolutions Picture 1 collected from “1983 Industrial Robots KUKA IR160/60, 601/60”

5

1.Cyber-physical systems

• Definition of Cyber-Physical systems (CPS) (Lee, Bagheri, & Kao, 2015): “Transformative technologiesfor managing interconnected systems between its physical assets and computational capabilities.”

ICT systems: Maintenance scheduled on time intervals

Virtual world

Equipment: Visual inspections, manually registered before sent to the virtual world.

Physical world

CPS:Convergence of the physical and virtual world

6

1.Cyber-physical systems Example of an architecture for CPS:

Figure 2: Architecture model for predictive maintenance (Source: Rødseth et. al., Euromaintenance 2016)

7

2. Smart Maintenance• German Standardization Report for Industry 4.0 (DIN, 2016): “In Industry 4.0 in general, and specifically in the factory of the future – the smart factory – maintenance will play a central role as the guarantor of the availability and reliability of machines and systems... Without systematic development of maintenance into smart maintenance, the successful implementation of Industry 4.0 will be put at risk.”

• Research priorities from the EU project Focus (www.focusonfof.eu): Optimized & Predictive Maintenance.

• Maintenance in digitalized manufacturing (Bokrantz et. al., 2017): Fact based maintenance planning.

8

2. Smart Maintenance

Maintenance 4.0

Figure 3: Industry 4.0 maturity model, adapted from (Nienke et.al., 2017) and (Schuh et. al., 2017)

9

2. Smart MaintenanceTable 1: The main value drivers for Industrie 4.0 adapted from (McKinsey&Company, 2015).

Value drivers

Service/aftersales Asset utilization

Activities and technology

- Remote maintenance- Predictive maintenance

- Remote monitoring and control

- Predictive maintenance

- Augmented reality for MRO

Indicative impact

10-40 % reduction of maintenance costs

30-50% reduction of total machine downtime

10

2. Smart Maintenance• Predictive maintenance (En 13306, 2010): “Condition based

maintenance carried out following a forecast derived from repeated analysis or known characteristics and evaluation of the significant parameters of the degradation of the item”

Maintenance

Preventive Maintenance

Condition Based Maintenance:

Scheduled, on request or continuous

Predetermined Maintenance

Scheduled

Corrective Maintenance

Deferred

Immediate

11

2. Smart Maintenance

Condition-based maintenance

DiagnosisPrognosis (Predictive

maintenance)

Physical-based Experience-based Data-driven

Physical-based:- Vibration analysis- Magnetism monitoring- Diesel Engine modelling- Tribology analysis- Process parameters analysis

Experience-based:- Performance monitoring of engine- Visual inspection

Data-driven:- Artificial Neural Networks (ANN)- Statistical models, e.g. Statistical Process Control

Source: Asmai, S.A. et. al (2010) 2nd International Conference on Computer Research and Development

12

2. Smart Maintenance

• Machine learning teaches computers where several algorithms “learn” directly from data.

• Evolvement from Artificial Intelligence.• Data is king with Moores Law.

13

2. Smart MaintenanceIn computerized maintenance management system (CMMS):Planned repair process, e.g. in SAP:1. Notification: Technical object, date, description, priority2. Planning: Work to be performed, material, tools, resource

internal/external3. Controlling: Order release, capacity leveling, paper printout,

availability check4. Implementation: Material withdrawal, external procurement5. Completion: Time confirmation, technical completion & confirmation

Record of history: Material usage, orders, notifications, information system, usage list

14

2. Smart MaintenanceWhat time window is needed in the maintenance system?

0

50

100

150

200

250

300

1 2 3 4 5 6 7 8 9 10

Capacity overview in CMMS, e.g in SAP

Required man-hours Overload Available man-hours Week

Man-hours

15

3. Application of Smart maintenance in cyber-physical systems

Figure 4: PLI calculations

Source: Rødseth, H. and P. Schjølberg (2016). “Data-driven Predictive Maintenance for Green Manufacturing.” Advanced Manufacturing and Automation VI, Atlantis Press. 24: 36-41.

16

3. Application of Smart maintenance in cyber-physical systems

Source: Rødseth, H., P. Schjølberg and A. Marhaug (2017). "Deep digital maintenance." Advances in Manufacturing 5(4): 299-310.

17

3. Application of Smart maintenance in cyber-physical systems

Source: Rødseth, H., P. Schjølberg and A. Marhaug (2017). "Deep digital maintenance." Advances in Manufacturing 5(4): 299-310.

18

4. Summary & conclusion

• Smart Maintenance applied in land based industry has been presented with a case of predictive maintenance.

• The benefit is to perform maintenance when actually needed.• A challenge is to have sufficient relevant data for machine learning.• To Transfer this Smart Maintenance into the maritime sector several

issues must be reviewed:– The cost of 128 kbs bandwidth as satellite service (MUNIN vessel) for notifying a future

maintenance action at deep sea. This service is a shared cost. Should this bandwidth be increased? How much data is actually necessary to transfer from vessel to shore?

– Is the machinery equipped with sensors measuring sufficient data quality? – Big data analytics on device (at the vessel) or in cloud (at shore)? – To what extent has the maritime industry willing to share anonymous fleet data?– If Shipping 4.0 permits manning of vessel, which maintenance tasks can be performed during

voyage?

19

The End

“Coming together is a beginning, staying together is progress, and working together is success.”

-Henry Ford-

Thank you for your attention!

Reference of open access publicationRødseth, H., P. Schjølberg and A. Marhaug (2017). "Deep digital maintenance.“Advances in Manufacturing 5(4): 299-310.

Available at: https://doi.org/10.1007/s40436-017-0202-9